Quality check and calculation of drought indices for the hydrological stations from the ADO project
Summary of Station ADO_DSC_SI03_0033 in Slovenia in HE Dravograd.
This dataset is aggregate at daily resolution with units in m3/s:
| country | region | location_site | lat | lon | start_date | end_date | waterourse | height_masl | height_dem | catchm_area | source_id | note_influence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Slovenia | Vzhodna Slovenija | HE Dravograd | 46.586959 | 15.023492 | 1965-01-01 | 2019-12-31 | Drava | None | 334.9377 | None | None | None |
Basic description and statistics of the dataset:
| ADO_DSC_SI03_0033 | |
|---|---|
| count | 20085.000 |
| mean | 260.717 |
| std | 139.993 |
| min | 57.852 |
| 25% | 159.531 |
| 50% | 227.999 |
| 75% | 322.000 |
| max | 1799.000 |
Description of periods with missing values in the dataset
| Start date | End date | Missing values |
|---|---|---|
| 1952-10-01 | 1964-12-31 | 4475 |
| 1984-12-31 | 1984-12-31 | 1 |
| 1988-12-31 | 1988-12-31 | 1 |
| 2004-12-31 | 2004-12-31 | 1 |
Dates exceeding the maximun threshold in the dataset defined as eleven times the standard deviation
| date | ADO_DSC_SI03_0033 |
|---|---|
| 1966-08-19 | 1789.0 |
| 1966-08-20 | 1799.0 |
Top Ten lowest values in the historical period
| date | ADO_DSC_SI03_0033 |
|---|---|
| 2002-01-01 | 57.85 |
| 2002-04-01 | 58.16 |
| 2002-03-31 | 60.38 |
| 2019-01-27 | 64.65 |
| 2006-12-06 | 64.93 |
| 2001-12-30 | 66.29 |
| 2002-01-24 | 69.17 |
| 1996-03-03 | 69.84 |
| 2003-04-12 | 70.09 |
| 2008-02-17 | 70.71 |
Top ten highest values in the historical period
| date | ADO_DSC_SI03_0033 |
|---|---|
| 1966-08-20 | 1799.00 |
| 1966-08-19 | 1789.00 |
| 1966-11-06 | 1670.00 |
| 1965-09-04 | 1669.00 |
| 1965-09-05 | 1632.00 |
| 1966-11-05 | 1584.00 |
| 2012-11-05 | 1416.98 |
| 1965-09-03 | 1411.00 |
| 1966-08-18 | 1399.00 |
| 1998-10-08 | 1375.82 |
Monthly discharge for the station ADO_DSC_SI03_0033
Hydrologic regime means the entire state of the streamflow in a given area over the course of a year. It is a function of the climate. The factors affecting the river regime include precipitation, temperature, evapotranspiration and the physical properties of the drainage basin. The following graphs describes the discharge regime for the multi-year period of records at different temporal scales. The shaded area represents the quantile 5 and 95.
Q95 is also called Q347. Is one of the low flow duration indices, denotes the discharge of the watercourse which is at least reached or exceeded for 95% of the time, that is on 347 days in an average year. For this station the Q95 value is 107.0 m3/s.
This index has been developed by Nalbantis and Tsakiris. Follow the basic concept of the Standardized Precipitation Index (SPI). The SDI can be computed at different time scales, where monthly streamflow is aggregated over a period (1-month, 3-months, 6-months, 9-months, 12-months).
The 1-month SDI is a short term index. Compares the streamflow of each month with the average streamflow over the historical record of the same month.
The 3-month SDI index provides a comparison of the streamflow over a specific 3-month period with the mean streamflow from the same 3-month period for all the years included in the historical record. A 3-month SDI gives a short and medium-term condition, and reflects a seasonal estimation of the streamflow condition. In this analysis is used the last quarter of the year to indicate the begining of the hydrological year:
The 6-month SDI provides a comparison of the streamflow over 6-month period with the average streamflow from the same period over the historical record. This index indicates a medium-term condition in streamflow. In this case it is used the first six months of the hydrological year.
The 12-month SDI provides an indication for the 12 consecutive months of the hydrological year with the same 12 consecutive months during all the years of the historical record. At this time scale the index reflects long-term streamflow patterns. Since, this time scale is the cumulative result of shorter periods. The longer SDI values tend to zero unless a specific trend is taking place. The Hydrological year goes from October to September.
In order to calculate the the Standardized Streamflow Index (SSI/SSFI) the following distributions are evaluated and their respected parameters are reported below. For the charts below it is considered the General Extreme Value (GEV) for the fitting in order to compare all the stations with the same approach. For more information on the considered distribution see index factsheet.
| sumsquare_error | Akaike information criterion | Bayesian information criterion | Kullback-Leibler divergence | |
|---|---|---|---|---|
| gamma | 0.000036 | 1459.806343 | -11011.653081 | inf |
| pearson3 | 0.000036 | 1459.806286 | -11011.653063 | inf |
| lognorm | 0.000043 | 1457.119164 | -10897.763904 | inf |
| genextreme | 0.000048 | 1463.607651 | -10824.272050 | inf |
| genpareto | 0.000082 | 1442.851792 | -10471.214113 | inf |
This test is used to decide if a sample comes from a population with a specific distribution, in this case if the fitted data comes from the General Extreme Value distribution. KS test returns two parameters D statistic and p-value, which D-statisctics is the absolute maximum distance between the two CDFs of the two samples, the closer to 0 the more likely is that the two samples belong to the same distribution. P-value corresponds to the probability value, in which the null hypothesis is rejected if the p-value is less than the significance level, in this case 0.05.
| General extreme value | D-statistic | P-value |
|---|---|---|
| Kolmogorov-Smirnov test | 0.044325 | 0.145075 |
It follows also the theoretical concept of the standardized precipitation index (SPI). Can be computed at different time scales, where monthly streamfloe is aggregated over a period of time, as the same as SDI. This index allows accurate spatial and temporal comparisons of the hydrological conditions of a stream (Telesca, 2011). This index compares normal streamflow for a specific rolling windows sum. For more information see index factsheet.
The 1-month SSI is a short term index. It compares normal streamflow for the month with the average streamflow over the historical data of the same month. The long-term streamflow record is fitted to a probability distribution. So the median SDI is zero. That means positive SDI values indicate greater than the median streamflow, and negative values indicate less than the median streamflow.
The 3-month is the time step for the rolling window sum.
The 6-month is the time step for the rolling window sum.
The 12-month is the time step for the rolling window sum.